25 August 2012

A Clash of Styles

After our first test of Non-random Fischer Random, which I documented in The Barbecue Positions, HarryO and I immediately started a second test. This time I had White.

I placed the first Bishop on the d-file, because I wanted to play with a center Bishop. HarryO placed the second Bishop on the e-file, because 'I really enjoyed the Bishops in the center in the past'. Whereas in our first game, HarryO had chosen to place the Queen for his second turn as White, I chose to place the two Knights. I put them on the c- and f-files to create a symmetrical position on the central files: **NBBN**.

For his second move, HarryO decided he wanted his King on the g-file, because 'I want to be able to castle short as soon as possible'. This left two squares for the Queen: either the a- or the b-file. He placed it on the a-file, because 'as Black, I don't want to add complications but want to simplify'. I imagine that would be a common reason for placing a Queen in the corner, especially against a good player. It takes the Queen longer to get into the game when it starts in the corner. Those starting choices gave us SP393 QRNBBNKR.

As White, I had the privilege of the first move and played 1.d4. Harry answered with 1...b5. That gave us the position shown in the diagram.


Some time ago, in Attention to the Chess960 Center, I observed,

There are two distinct, fundamental ways to treat a chess960 opening. The first way is to follow traditional chess opening principles, of which one of the most important is to pay attention to the center. The second way is to pay less attention to the center, but by taking into account the specific start position, to emphasize the rapid development of the pieces to good squares, even if this means making early moves like g4 or b4.

I am definitely in the first camp, following traditional chess opening principles. After playing a few games with HarryO, I now know that he is in the second camp, paying less attention to the center. That makes our games a clash of styles from the outset.

I spent a lot of time studying the position after 1...b5, looked at many different moves that adhered to classical principles, and finally decided that the non-traditional 2.a4 was my best shot. It solved the problem of developing my Queen and gave Black an immediate problem.

The rest of the moves and associated commentary -- we are only playing to move ten in these trial games -- can be found on HarryO's blog, Non-Random Chess960 Trial Game 2. Our first two non-random chess960 games have shown that the players can indeed determine the start position without any special equipment. Whether it is the best way to do so remains to be seen.


GeneM said...

Mark wrote:
follow traditional chess opening principles, of which one of the most important is to pay attention to the center

decided that the non-traditional 2.a4 was my best shot

I respectfully caution us against the casual habit of using the powerful word "principles" when describing chess opening theory that is based on deep experience with only the traditional start setup. I believe some of what we today call opening "principles" will eventually be exposed as being mere esoteric tactical considerations of the traditional setup.

Your non-traditional early opening move 2. a24 might well obey an important opening priciple of pure chess, even if it violates what we all have sloppily or ignorantly been calling an opening principle based on our very narrow experience limited to only one start setup.

Until humanity settles on exactly one second start setup to reuse for several years and a million games, humanity's opinion of what the opening principles are for pure chess is highly suspect --- the say published analysis of moves and games from the 1980's was later found by Fritz to have been wrong all along.

Discard the 'Random' from Fischer Random Chess!

Thank you,

Mark Weeks said...

GeneM - I've discussed that subject, which I sometimes call meta-theory, in previous posts...

Knights before Bishops?

Lasker on Basic Opening Principles

Yusupov's 'General Principles of Opening Play'

Fine's 'General Principles' of Opening Theory

...You write, 'Until humanity settles on exactly one second start setup to reuse [...], humanity's opinion of what the opening principles are for pure chess is highly suspect'. How do you reckon that studying a single new start position will provide more insight than studying all 959 new start positions? You risk falling into the same rut that we have for the traditional start position. To understand the universe, which is better: studying a single star -or- studying all stars?

As for 'your non-traditional early opening move 2.a4 might well obey an important opening priciple of pure chess', I believe that it does indeed. The subject, however, goes too far beyond the point I wanted to make in this current post. Watch this space. - Mark

GeneM said...

Mark replied with:
How do you reckon that studying a single new start position will provide more insight than studying all 959 new start positions?

I so reckon because nobody will 959 other endlessly randomly chosen start setups in the kind of depth necessary to truly compare their opening trees to the traditional opening tree (which has been deeply honed over centuries).

The best you could hope for is that by the year 2050 both computer hardware and chess engine software will have advance so much that a bank of affordable computers could generate an MCO for a chosen second setup.
But then there is a big practical problem --- Who would care about looking at it and talking about it --- unless it is a second stable setup that is frequently reused on real tournaments or matches?: nobody would.

Morphy said...

Chess 960 is a popular variant because of the random start position.
Choosing one position from chess 960 to be replayed all the time is a different variant which has zero popularity (compared to chess 960) and is in my opinion inferior to the following simpler variant ALREADY tried out: Displacement Chess organized in 1935 by P.Keres.
Simply switch the king and queen for either black or white. Keep Castling simple with king moving two squares to left or right. Now you have a beautiful start position (not an ugly chess 960 one) and brand new opening theory -- better than any one chess960 position.

- Note that the position in Fischer Random with king and queen switched with FRCastling will feel more alien to a regular chess player than the king and q switched on one side with regular 2 step king castling.

see this site for a modification of that variant by adding "free" castling rules.

HarryO said...

Morphy, thanks for the link to Displacement chess. I was initially excited by it. Sometimes I have to pour a bucket of cold water over my head with my personal love of Chess960...

The big attraction to displacement chess is that it encourages short side castling to opposite sides of the board. That issue alone could be a serious problem for black. If that isn't a problem, then it probably seriously skews the available strategies into a narrow band than traditional chess because of that one single change to the position of the kings...

Therefore, I doubt that displacement chess is sufficiently "equalized" for both sides.

You could change displacement chess so that the king's are on the same files, but then you have the problem that there is probably very little point to castle into the traditional chess positions, when it is basically always better for white to castle with rooks onto the central files.

*I'm only guessing* that this is the reason why it was dropped.

Ironically, I think that it is possible that Bobby Fischer's invention FRC (Chess960) is probably more equalized overall than displacement chess is, when you look across all 960 positions and do not get tempted into myopically isolating out the corner cases that are a bit more tricky for black.

That said, displacement chess is a pretty good effort at an attempt to improve on chess and because my opinion is essentially irrelevant, people might like it still.


Morphy said...

I don't think neither of us can say whether displacement chess is any less equalized than some of the chess960 positions.
Sure it encourages opposite wing castling but when talented players find a way for white to obtain an advantage others will find a way for black to equalize. Black does not have to castle opposite wing. Is some standard chess openings black must wait for white to castle before castling on same side or face a mounting attack. Same principle here.
On the other hand chess 960 positions feel less natural simply because of the awkwardness of stepping on one's own toes so to speak. If one just wants a new position to start from there is no reason to pick one of them and i feel displacement chess is a better alternative.
On the other hand, I love the fact that you get a different start position with chess 960 _ and i think this is the form that will gain popularity. Forcing players to play one position picked by some guy who thinks he knows better seems silly and pointless esp when other alternatives exist.
IN fact the random start position sort of explains the jumbled weird position you get sometimes.